Global fixed income markets are undergoing fundamental changes due to the growth in credit markets and a fast dwindling supply of sovereign paper. In turn, investors are increasingly needing to address the question of whether traditional benchmarks are suitable measures of performance.
Index providers are convinced that the credit indices they have been rolling out since 1997, when European broad market benchmarks made their first appearance, represent the future of fixed income investing.
Few would disagree with them. Bond markets are undergoing fundamental change, and although the majority of investors outside the US market are new to the notion of credit, the reallocation of European portfolios towards spread product and the broader global realignment of fixed income markets towards non-sovereign debt looks irreversible. In turn, the suitability of domestic and government indices as portfolio benchmarks is a question that is becoming increasingly hard to ignore.
In Europe in particular, the introduction of the euro and the wider backdrop of more intensely competitive global business markets is leading to explosive growth in spread product, with corporate issuance leading the way.
Indeed, rarely have different forces conspired to bring about the rapid growth of a marketplace. On the one hand, fixed income investors are having to respond to the removal of value opportunities in European sovereign markets due to yield curve convergence and the disappearance of 10 currencies.
On the other hand, the need for financing among corporates has increased dramatically as industries consolidate and restructure.
Bank disintermediation is contributing to the resulting transformation of the European corporate bond market from underdeveloped backwater to mainstream funding source, with banks focusing increasingly on return on equity and corporates for their part recognising that bank capital is becoming a less certain source of borrowing.
But although the euro's arrival is the unique catalyst behind the growth in credit markets in Europe, other factors are driving a realignment of fixed income portfolios towards spread product in markets globally.
Investors have been faced not only with diminishing returns in sovereign markets due to a sharply lower global interest rate environment - sovereign supply is also dwindling across many markets. The UK, for instance, provides a textbook case of a market that is witnessing an evaporation in sovereign supply as well as strong growth in corporate bond issuance.
As Barclays Capital noted when launching its new Sterling Bond Index in January 2000: "The diminishing borrowing need of the UK government is leading to a relative shortage of Gilts, while the sterling corporate bond sector is growing rapidly. Government bonds accounted for less than 25% of new sterling issues in 1999, compared to 72% in 1992. Of the £58bn raised in 1999, 77% came from non-UK government issuers."
The same pattern of reduced government financing is repeated across many of the major markets outside Japan.
Illustrating the trend, Merrill Lynch noted in its 1999 Bond Index Almanac that only 19 countries in the bank's Global Government Index extended duration that year. "The trend in 1999 was clearly toward reduced government financing," the report said, "which, along with a rising yield curve environment, was a major contributor to the shortening trend."
The report continued: "The face value of most country indices was reduced during 1999 - some substantially. Only Japan (+12%) and the euro index (+11%) experienced a sizeable increase in total amount outstanding."
Among the euro sovereign group, though, Italy alone accounted for Eu87bn of the Eu205bn net increase in face value, and the overall trend in the euro area is towards reductions in debt outstanding. Research from JP Morgan's European fixed income division, for example, estimates that gross bond issuance in the euro area will be down 91.1% in 2000 versus 1999 - partly as a result of revenues from the sale of UMTS licences.
With spread product expanding rapidly in the other direction, the share of outstanding sovereign bond issuance in Euroland has already fallen from almost 79% in 1996 to 61% in 1999.
Reduced government financing has been most marked in the US. In 1999, for example, decreased federal financing reduced the Treasury weight in Merrill Lynch's US Domestic Master Index by 4% - bringing Treasury allocation below 40% for the first time since 1983.
On the opposite side of the globe, but sharing the same trend, the government sector's share of Merrill Lynch's Australian Composite Index fell 1% in 1999, due to a A$6bn decrease in face value and a substantially lower rate of return versus semi-governments.
Moreover, echoing trends elsewhere, Australia has seen a dramatic expansion in the depth and scope of its credit market over the last 12-18 months, with the share of its corporate market rising from 16.9% of the total fixed income sector in 1995 to 26% at the end of 1997 and to almost 36% by the start of 2000.
"The most notable development in the Australian dollar debt market is happening in the corporate sector of the domestic market," noted Merrill Lynch in its 2000 Guide to the Size and Structure of the World Bond Market. "Until recently, corporate issuers have largely underestimated the Australian marketplace, as they were able to find much more favourable funding conditions in the developed markets of Europe, the US and Japan, or to finance their operations through bank loans.
"Concurrently, Australian institutional investors traditionally concentrated their investment planning on national government or state government issues. The current condition of Australia's political economy, however, points away from this polarity and toward the creation of a vibrant credit market place, marrying a diversified group of issuers and investors."
In 2000, the Australian corporate bond market continued to broaden and become more diversified in terms of ratings and duration. When triple-B rated Southcorp launched a A$100m 10 year issue via Westpac, ANZ and CBA in March, for instance, it established an entirely new benchmark for BBB credits at the longer end of the yield curve.
Inevitably, developments of this kind will bring new demand on the part of local as well as international investors for performance yardsticks that accurately reflect the growing importance of corporate bonds in the Australian market.
Corporate issuance is also increasing across other markets in the Asia Pacific Region (APR), although this is happening in absolute rather than relative terms. Hong Kong's government, for example, issued virtually no debt in 1991, when government paper accounted for 0.5% of the total Hong Kong dollar bond market. This share has since risen to 10% of the total, with the overall share of the corporate market falling in direct proportion. Nonetheless, in absolute terms, issuance in the Hong Kong market has soared over the last decade, from HK$23.2bn in 1991 to HK$308bn outstanding at the end of 1999.
For index providers including Lehman Brothers, Salomon Smith Barney and Merrill Lynch, these changes have led first to the roll-out of a full suite of regional broad market indices, and second to the creation of global credit indices based on these regional building blocks. In the case of global broad market indices, Lehman Brothers was first out the blocks - albeit with an APR weight which initially consisted entirely of government and related debt - with the launch of its Global Aggregate Index at the beginning of 2000. And like Merrill Lynch, Salomon Smith Barney unveiled a global credit index - WorldBIG - later in the year.
In the case of the Merrill roll-out, an Australian Broad Market index, a Japan Broad Market index and an Asian dollar Broad Market benchmark were all introduced in the months running up to the launch of the global broad market offering.
Under the umbrella of "global" and "broad", Merrill has developed three distinct series of indices. These are the Global Broad Market series, with over 14,000 investment grade securities from the world's major bond markets and with a total capitalisation of over $13tr, in which the cut off size is $150m/ Eu100m for non-sovereigns and $1bn/Eu1bn for sovereigns; the Global Large Cap series, in which the cut off size for non-sovereigns is raised to $500m/ Eu500m, meaning that although its capitalisation falls to just under $11tr the number of bonds included in the index falls to under 4000; and the Global Broad Market Plus, which adds a number of smaller bond markets.
Like Merrill, Lehman Brothers has opted for a series of global indices based on different liquidity thresholds. That strategy led Lehman to increase the cut off size for non-sovereign issues in the Global Aggregate from $150m to $300m/Eu300m, with the changes effective from October.
"We know that we leave a pretty large footprint and can have a very significant impact on peoples' investment strategy," says Steve Berkley, head of global fixed income indices at Lehman Brothers in New York, "so we want to make absolutely certain that we are doing the right thing in terms of liquidity."
"A number of investors wanted a roll up index, a number wanted a $300m cut off point and some wanted an even higher liquidity constraint. But we found that if we raised the cut off to, say, $500m we would have ended up with mostly a government bond index."
Nonetheless, Lehman is planning to introduce a $500m cut off in response to investor demand, and also has in the pipeline what Berkley calls the "kitchen sink index", which will incorporate all asset classes regardless of currency or investment quality.
Global broad market benchmarks including many thousands of bonds are a new departure for investors, and Philip Galdi, managing director of quantitative analysis and portfolio strategy at Merrill Lynch in New York, concedes that the transition from a government index to an index such as the Global Broad Market can seem a daunting task for fund managers. But he says the apparent predicament for investors arises more from perception than reality.
"These new indices are going to be used by global fixed income fund managers whose traditional exposure has been largely to sovereigns," he says. "That means that they have to make a huge leap of faith in moving from an index including, say, 600 to 700 bonds to one with 14,000. I can sit down with an investor and explain to him how many bonds he would actually need in order to replicate the index with a neutral position - which is not many. Nevertheless, I accept that 14,000 bonds can sound like a very large and daunting number."
Drilling down into the way an index is composed, however, underscores Galdi's point about replicating a much broader benchmark. As an example, he notes that "Merrill's broadest index includes about 14,000 bonds, of which about 3000 are German Pfandbriefe. Given the nature of these securities, the 3000 can easily be replicated with only a handful of representative issues."
For some investors, the large number of bonds included in a global index is far from daunting. Accompanying the Merrill product at its launch was the Broad Market Plus benchmark, which added a range of smaller and emerging markets such as Poland, Hungary, the Czech Republic, Taiwan, Singapore and others to the index. Galdi says the index was in response to the sheer diversity of investor opinion on whether or not to include smaller markets within a broad global index.
"Prior to launching the index we asked investors whether they wanted the smaller markets included or cut back," he says. "And most saw value in keeping the smaller bits and pieces out of the index. But a number of investors commented to me that some of the most unique investment opportunities and compelling value was to be found in these smaller markets." Many of those investors, Galdi adds, will also have policy guidelines within their mandates that allow them to invest only in markets nominated within their official benchmark. In other words, if these investors want to pursue a strategy of, say, overweighting Poland or South Africa within a global portfolio, they can do so only if they use a benchmark which includes these markets in the first place.
The long process of introducing regional broad market indices before pulling them all together in a unified global package has naturally presented index providers with considerable challenges - including the need to assess local conventions against global standards. Often the two can conflict - for instance, when it comes to the issue of ratings, which have historically been much less broadly accepted by issuers (or demanded by investors) in Europe or Asia than they are in the US.
This meant that an early consideration for index providers with their roots in the US was whether or not to include non-rated bonds in their benchmarks. "When we were developing our sterling index," says Galdi at Merrill Lynch, "one of the key areas we had to address was ratings requirements. We decided to include non-rated bonds in our sterling All-Stocks benchmark which was something we would never have done in the US market. But there are a reasonable number of unrated Sterling issues and a number of these are bonds that many UK investors would not dream of leaving out of an index. At the same time we offer a Sterling Broad Market index that abides by our global rating standards and allows for consistency within our Global Broad Market series."
The question of ratings also creates challenges in a Japanese context, albeit for different reasons. The problem here, says Galdi, is the sharp divergence between the ratings assigned to Japanese issuers by local agencies in comparison with those assigned by Moody's or Standard & Poor's.
International ratings agencies will often rate Japanese issuers two or three notches below the domestic agencies, again creating significant differences in the way local and international investors appraise the Japanese bond market. "So in Japan we also created parallel indices," says Galdi. "The Japan broad market index follows our global standards using Moody's and S&P, whereas the Japan Extended benchmark looks more like existing domestic indices." The Japan Extended Index still takes Moody's and S&P as its principal arbiters of credit-standing, but also allows for the inclusion of bonds that are not rated by these agencies, but which have a minimum rating of A- from R&I or JCR.
An important reason for the emergence of new global benchmarks - besides reduced sovereign supply in many markets and the growth of credit product - is Japan's increasing share of the global government bond market, due to the accelerating slide of Japan's public finances into the red.
Not surprisingly, Japan's weight in global government bond indices has soared - the weight will exceed 30% of Lehman Brothers' global government bond index, for instance, in 2000. For global fund managers this is clearly a problem, since they must either buy more JGBs simply to stand still, or else add further tracking error to their benchmarks.
Most global fund managers have in any case tended to be significantly underweight Japan, reflecting not only reluctance towards having a disproportionate amount of JGB paper in the portfolio, but also caution over being in a market which by global standards is extremely illiquid.
This, say index providers, is precisely why using global credit indices as a benchmark may now be the most astute way to manage exposure to the Japanese fixed income market.
"A clear problem that government bond index managers are facing at the moment is a benchmark that is going to become ever more dominated by JGBs," says Berkley at Lehman Brothers in New York. "So by including Japan on the global stage within a credit universe you can greatly reduce Japan's overall share in your benchmark."
Compared to Japan weightings of 25%-30% or more in global government bond indices, the weight in global credit indices is around 15% - quite literally a weight off international fund managers' shoulders.
Although the changing composition of fixed income markets is clearly promoting demand for more representative performance measurement benchmarks among institutional investors worldwide, there is a division of opinion about how far this process has advanced from market to market.
In the context of Euroland, it is not difficult to identify some of the frontrunners as the investor base starts shifting from use of national, government-oriented benchmarks to more internationally diversified indices. Index providers cite large German institutions such as DWS and Deka and leading Dutch pension funds such as ABP and PGGM as examples of institutions that have revised their benchmarks.
But while the trend towards use of broader based benchmarks is self-evident, some say that it should not be overstated - at least, not at this stage in the evolution of the credit markets.
Steven Bayly, Frankfurt-based head of the Euroland portfolio management team at Dresdner Investment Trust (DIT) - the investment management division of Dresdner Bank - is keen to emphasise that government-based indices still play a critical role for his clients.
"Of course, credit has been the growth area within the fixed income market," he says, "and we are seeing more and more institutional clients demanding dedicated credit portfolios. But in terms of absolute size, credit is still not a massive part of our clients' traditional portfolios.
"Generally, traditional government benchmarks are still sold to clients, although a clear trend can be seen towards more differentiated indices which include credit."
DIT's Bayly acknowledges, however, that there are disadvantages associated with sticking to Treasury only benchmarks - especially for mixed portfolios that are inching away from having a government bias.
"If you have a negative view on credit spreads you cannot really portray that if your index is 100% government-based because you cannot go short," he says.
By using a government index, in other words, to measure a portfolio with a growing credit element, are investors not exposed to being caught short when swap spreads widen, as they have done so dramatically in 2000?
"Sure," says Bayly. "That is always the danger. In our case it has not been so much of a problem because we threw out most of our long dated non-government paper at the start of the year. In the shorter dated paper, it is less of an issue because we can basically hold the bonds to maturity and lock in a certain spread."
At Schroder Salomon Smith Barney in London, director of fixed income research Rafey Sayood confirms that in spite of the rising importance of credit, the continued importance of government bond exposure among European institutions should not be underestimated.
"There is still a large number of fund managers in Europe and Asia that have a portfolio position that is essentially government-based, with the proportion invested in credits ranging from, say, 5% to 30%," he says.
Others also point out to the continuing importance of traditional benchmarks for institutions in Europe. Speaking in a workshop session at a recent Pfandbrief conference in London, for example, Paola Lamedica, credit bond strategist at BNP Paribas, suggested that the credit culture has barely scratched the surface in continental Europe.
"When we speak to fixed income investors, especially in continental Europe, we find that they are still analysing and studying credit indices," she said, "but relatively few of them have actually made the move out of government indices and into credit indices."
Some of that might come down to the mind-boggling array of indices that investors are now being asked to assess or compare - reviewing the relative advantages of the product being offered should not be rushed - but in part it may also come down to simple inertia. A recent report on credit indices from Lehman Brothers referred to the "index complacency or outright indifference among some European bond managers", but also predicted that it is only a matter of time before inertia is "swept away by the riptides of global economic, capital market and asset management methodological convergence for European bond managers".
In fairness, there are good reasons why fund managers are taking their time to switch from government benchmarks to broader credit indices. The euro credit market is in its infancy, with the government sector accounting for between 60% to 80% of euro area broad market benchmarks. In contrast, depending on how the index is structured, corporate bonds and utility paper accounts for less than 3% to 5% of the indices.
And with the share of triple-A and double-A rated paper accounting for more than 90% of these indices, the scope for investors to move down the credit curve remains limited.
Over time, of course, that will change. In Merrill Lynch's EMU Broad Market, for instance, the non-financial sector saw the fastest growth in 1999 - increasing by almost 2% to 3.4% of the index.
But whether it is inertia or a more understandable response to the current structure of the markets, institutional investors throughout Europe will have to face up to the challenge of working out which index suits their investment needs most closely. And identifying which index to run with is not a straightforward process, even for the most experienced firms of consultants.
"What was especially difficult from my perspective nine or 12 months ago was that there was a massive race among some of the banks to be the index provider of choice," says Robert Marsden, a consultant at Mercer & Co in London. "Developments were unfolding so rapidly that we were very reluctant to pin ourselves down to one provider that could quite quickly have been superceded by another."
For its part, Mercer has now settled on a basket of three providers in the UK and four on a global basis that it is comfortable with recommending to its clients. Although Marsden chooses not to identify the names on this short list, rocket science is not required to figure out which they are.
Some observers question the assertion that the trend towards indexation has yet to filter through to a broad cross-section of European institutions.
"In my travels across Europe over the last year or so I have visited some very small accounts in countries such as Luxembourg or Germany that are active users of credit and high yield indices," says Preston Peacock, vice president, quantitative analysis and portfolio strategy at Merrill Lynch in London. "We are talking about a phenomenon here that is certainly no longer confined to the largest players in the market."
And if some European institutions have been slow to adopt credit indices, all the evidence suggests there is no question that investors have vastly increased their exposure to credit products since the launch of the euro at the start of 1999.
"In 1998 when we were selling credit products there were basically two countries in Europe we could go to - the UK and France," said Manfred Schepers, managing director and global head of debt capital markets at UBS Warburg. "Two years later that has dramatically changed. We have seen countries like Germany and the Netherlands become much more active in the purchase of credit."
That development, he added, has had a clear knock-on impact in terms of roadshows. "Roadshows are now taking issuers to all financial centres - to cities like Stockholm which in the past were never visited by issuers. Nowadays about 15% of our corporate distribution is typically in Scandinavia, whereas two years ago that would have been less than 1%."
But if there is still a mismatch between the extent to which investors in Euroland buy credit, and the extent to which they actually use credit indices, that can only be good news for providers of credit indices targeting European institutions as their prospective clients. And there is no doubting that providers are bullish about the prospects for increased use of indexation by the Europe investor base.
For example, Lehman Brothers nailed its colours firmly to the mast on this subject in its recent report. "Within five years," this noted, "we predict the near impossibility of finding a major European steward of fixed income assets that does not rely heavily on a comprehensive fixed income index to measure performance and risk."
Whether or not the same could be said of Asia is a different story. Chris Francis is head of global credit research at Merrill Lynch in London, but transferred recently from Hong Kong, where he headed the bank's Asian fixed income research. He says that while Japan is following a similar pattern to Europe in terms of an expanding corporate bond market and accompanying benchmarking, elsewhere in Asia the use of credit indices is still a long way from having taken hold as a benchmarking tool.
"A lot of the buyers of credit in non-Japan Asia are banks, corporates and private client individuals," he says, "so there has been less institutionalisation of benchmarking there than there has been in Japan or Europe. I think that will change slowly as fund managers become more accountable, but at the moment our indices are used more for analytical purposes in non-Japan Asia than as benchmarking tools."
Credit risk/sovereign risk
Aside from the continuing expansion in the size of European and other corporate bond markets, there are a number of important and closely related forces at play which will drive increased use of credit indexes as barometers of performance.
Among these is the sudden and brutal emergence of event risk in the credit markets and the white-knuckle volatility that has accompanied it. This has given fund managers a painful wake-up call - alerting them to the fact that corporate bonds should not be seen as a convenient way of incorporating an attractive looking yield kicker into a fixed income portfolio at very little risk, which appeared to be a commonly held view in the early days of the credit markets.
There was a good reason for this. When corporate bonds were relatively insubstantial components of the overall fixed income markets in regions outside the US, they were used as a highly effective way of outperforming traditional benchmarks with a modest allocation of, say, 5% of corporate paper into a portfolio dominated by government paper.
This practice was prevalent until recently, for example, in the UK retail fund management industry. When demand in the UK for corporate bond PEPs and ISAs took off in latter half of the 1990s - accompanied by a marketing blitz among the leading retail fund managers - investment managers persisted in using demonstrably misleading benchmarks to track their performance. For obvious reasons, the preference was to market corporate bond funds by highlighting their outperformance relative to the government bond index, which was a comparison, par excellence, of apples and oranges.
"For a time it was perceived that the easiest way to beat the government benchmark was to use credit," says William Lloyd, head of fixed income indices at Barclays Capital in London, "because it allowed investors to capture wide spreads and if those spreads then tightened in the secondary market, so much the better.
"The real driver towards more solid credit benchmarks came at the end of 1998, when credit spreads suddenly gapped out and all of a sudden the strategy of using corporate bonds as a way of outperforming government benchmarks was not working any more. I think that led a lot of investors to make a very serious re-evaluation of what their benchmarks were."
It has certainly forced the UK retail fund management to adopt more defensive language when explaining the performance of its funds to its investor base. Take the latest Investment Review published by M&G's highly successful £1.1bn UK corporate bond fund, which has close to 90% of its assets under management invested in credit products. "Over the review period, the fund's total return was 2.5%," the investment review explained, "below the total return of 5% from the FTSE Gilts 5-15 Year Index. However this is not comparing like with like, as this index does not measure the performance of corporate bonds, only of UK Gilts which have outperformed due to technical factors." But there was no mention of comparing like with like in previous reviews from the same manager, when the fund was happily outperforming its government benchmark.
A related reason explaining the steady pick-up in demand for credit indices is investors' inexperience in dealing with corporate bonds as an asset class. With the presence of experienced credit analysts a rarity at continental European institutions, fund managers who think they can outperform their peers, let alone an objective benchmark, could be accused of a certain over-optimism. Recent spread widening in the market will (or should) have persuaded those with overly ambitious targets that a more sensible strategy, in the short term at least, would be to accept median returns by at least partially tracking indices in a more passive manner.
But even for those investors who may choose to track a credit benchmark passively, the vicious volatility of credit markets over the last year or so begs a very obvious question. If spreads have widened so dramatically recently and if credit is so difficult to analyse, are not providers of new credit indices behaving like farmers promoting Christmas Day to a community of turkeys? In other words, would not investors be better off avoiding this new asset class, and the indices that have accompanied its arrival, altogether? The answer over the medium to long term clearly has to be no, given the structural changes which point to the credit markets' ascendancy over the sovereign sector.
But Galdi at Merrill Lynch in New York, accepts that investors looking at the recent performance of credit markets might well resist advice to take a closer look at the asset - at least for the time being. "You cannot ignore the fact that the recent past does highlight that there is a downside to owning this asset class," he says, "and that the downside can be very significant."
When Merill Lynch launched its new flagship Global Broad Market Index (GBMI) in September, the bank did not shy away from addressing the issue of the credit underperformance lately. Far from it. Instead of looking simply at the recent performance of the asset, which Galdi says is "not enough of a baseline on which to judge the merits of the asset class", Merrill Lynch presented a comprehensive analysis of the performance of corporate bonds over the last 20 years.
The analysis demonstrated that over an extended period, credit (in the US market, at least) does outperform government and quasi-government paper. "Since it offers the deepest test bed of historical data with which to evaluate the longterm performance of credit," the research noted, "we have focused our attention on a 20 year analysis of incremental return performance in the US corporate bond market. "The results suggest that, yes, in most market environments, corporates will produce incremental returns in an amount commensurate with their average yield spread."
Specifically, according to the Merrill Lynch data, "when the 20 year results are broken down into rolling five year periods, it becomes apparent that the return premium for holding corporates has followed a relatively stable, albeit downward trending pattern - moving from about 150bp per year in the early 1980s to about 50bp per year in more recent times."
Some caveats are needed to this analysis. There is clearly a degree to which comparing the historical performance of the US credit market to likely developments in the European market is a comparison of different animals. In fairness, Merrill Lynch's research makes no pretence to the contrary. "When things go bad [in the credit market], they can go bad in a hurry and in a very big way - particularly in the case of the long duration segment of the corporate bond market," the report warned. "In other words, buyer beware!"
There is no doubting, though, that by rolling out credit indices, providers are not only offering investors new standards which will come into their own over time, but are also extending an important helping hand to fund managers as they take their first steps as investors in credit.
"Our new indices, together with the extensive analytics offered, provide a valuable resource of market data and analysis for investors who are new to this asset class," says Merrill Lynch's Galdi. "As credit begins to consume a larger portion of their holdings, the indices also serve as a more representative standard against which to measure performance." *
Japan's fixed income conundrum
Japan's spiralling debt burden is something that global fixed income investment managers would rather do without.
The problem is simple enough. If global managers are to replicate almost any global index faithfully and accurately, they will inevitably need to hold a large and fast growing share of Japanese paper in general, and of Japanese government bonds (JGBs) in particular. This is because, while governments in the US and Europe have been reducing their overall debt and in some instances buying back government bonds aggressively, JGB issuance has escalated almost uncontrollably in the wake of Japan's economic crisis in the 1990s. Since 1992, Japan has issued 10 comprehensive economic stimulus packages for a total sum of ¥140tr, and funding these outlays has seen the Japan weight in global indices rise inexorably.
According to Merrill Lynch data, Japan accounted for 18.2% of the world bond market at the end of 1999, with 50% of the total contributed by JGBs. In the global government index arena, the share of Japan in Merrill Lynch's index grew by what the bank describes as a "stunning" 6% over the course of 1999.
In that year, the high weight of Japan in global benchmarks worked in favour of index tracking bond fund managers - reflecting the outperformance of the Japanese market relative to other developed fixed income markets on the one hand, and the strength of the Japanese yen on the other. This has been especially relevant for investors in Euroland, given the weakness of the single European currency - the yen rose in value against the euro by 29% in 1999, compared with a rise against the dollar and sterling of 10%. As Richard Woodworth, global strategist and former chief Japan strategist at Merrill Lynch in London explains: "For European investors more than for US investors, underweighting Japan against a global government bond index has been the single most important source of underperformance over the last 12-18 months."
How long this can last is debatable, although Woodworth says that, for the time being, Merrill Lynch remains positive about the Japanese market - and is even recommending that clients maintain a marginally overweight position in a global portfolio. But he adds: "The current situation of exceptionally low yields in the JGB market is not sustainable. At some point the economy will have to recover and the deflationary forces that continue to be very much in evidence are going to recede. If Japan does break out of its current cycle, at some stage we will be talking about real interest rates of about 3% and inflation at 2%, which means bond yields at 5% - which is on a par with yields in other major markets."
If the tide were to turn relatively quickly, says Woodworth - with yields on JGBs spiking in short order from 2% to 5% - the impact on fixed income investors with market weightings in global government bond indices would be "devastating".
Moreover, the structure of the Japanese fixed income market would compound the problem. "Relative to other major government bond markets," says Woodsworth, "the JGB market remains illiquid. Only one issue trades with much liquidity and that is in the 10 year maturity. If the economy does start to normalise, a lot of investors are going to be anxious to get out of the door at the same time."
The problem, then, is that investors are in danger of being damned if they do and damned if they don't. If they continue to underweight their Japanese positions they face the risk of underperforming the global benchmark.
But if they maintain either a neutral weighting or a modest overweight position, they will potentially be sitting on a time bomb, due to the difficulties of exiting the market.
A partial way around the problem is for global bond fund managers to buy more liquid - (relatively speaking) and highly rated Euroyen and global yen instruments, and issuers have been only too happy to provide a steady stream of new issues to cater to this demand.
These include first time borrowers such as Austria's AAA-rated Oesterreichische Kontrollbank (OKB), which isued a 10 year global yen bond, guaranteed by the Austrian government, priced flat to the 2010 JGB no 219.
Like the OKB transaction, Deutsche Ausgleichsbank's (DtA) inaugural global yen transaction - a ¥100bn 10 year priced at 2bp over the JGB - was comfortably absorbed by investors eager to maintain representative weightings in the yen market while reducing dependence on JGBs.
But although playing the Euroyen market in preference to JGBs may help investors address the liquidity problem in the Japanese market, it will not go far towards resolving the issue of sitting on a large yen position which could quickly turn from being yesterday's star performer into tomorrow's painful underperformer.
It is partly for this reason that global credit indices have been introduced - giving investors a lower Japan weight versus global government bond indices.
Some investors, though, still complain that the structure of the Japanese market is such that it continues to present an intractable problem for index providers looking to produce global benchmarks with more or less uniform rules for inclusion, and that the solution would be to bring in entirely new rules for calculating what Japan's weight should be.
As one London-based investor says, size and liquidity are less correlated in the Japanese market than in any other. "There are some very large municipal and construction bonds outstanding in the Japanese market," he says, "which I guess need to be classified as credit. But they tend to be held by the local banks and virtually never trade.
So should they be included in global credit indices? Probably not, because one of the most basic requirements of an index is that the portfolio manager can replicate the benchmark. If you cannot replicate an index how can you hope to outperform it?"
"If you are looking for a market cap-based index on the one hand," says another global fund manager that is grappling with the Japan problem, "but if you also want rules to be applied consistently across all markets, Japan is certainly a conundrum. There is no perfect answer except for the Japanese government to issue fewer bonds. Unfortunately we cannot really send a delegation to Tokyo asking them to borrow less." *